Lexos Media IP, LLC v. Overstock.Com, Inc.

2:22-cv-02324Ksd02.02.2026

Gesamter Gesetzestext

IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF KANSAS
LEXOS MEDIA IP, LLC,
Plaintiff,
v.
OVERSTOCK.COM, INC.,
Defendant.
Case No. 22-2324-JAR
MEMORANDUM AND ORDER
Plaintiff Lexos Media IP, LLC (“Lexos”) brings this patent infringement action against
Defendant Overstock.com, Inc. (“Overstock”), alleging it infringed three patents that relate to
modifying the display of a cursor image on a website. This matter is before the Court on Lexos’s
Opposition to Defendant’s Motion to Exclude the Testimony of Samule [sic] Russ, Ph.D. (Doc.
194), and Response to Defendant’s Motion for Summary Judgment (Doc. 193). The Court
previously directed each attorney for Lexos listed on the signature blocks of Docs. 193 and 194
to show cause in writing why they should not be sanctioned under Fed. R. Civ. P. 11 and referred
to the relevant disciplinary administrators for signing and submitting documents in this case that
contain defective legal citations created through the use of generative artificial intelligence
(“AI”), which were not checked and confirmed before filing.
Lexos’s attorneys responded to the Court’s Order to Show Cause with declarations, and
the Court is prepared to rule on whether they should be sanctioned under Rule 11. As described
more fully below, the Court finds that all five Lexos attorneys violated Rule 11(b) by signing
filings that included fabricated and misrepresented legal authority. The Court releases counsel
Michael Doell from further sanctions. The Court sanctions the remaining attorneys
proportionate to their role in the violations.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 1 of 36
2
I. 	Background
A. 	Attorneys of Record for Lexos and the July 7, 2025 Filings
Six attorneys are listed as counsel of record for Lexos. Kenneth P. Kula, Michael W.
Doell, Christopher M. Joe, and Eric W. Buether from the Texas law firm Buether Joe &
Counselors, LLC are all admitted pro hac vice. Sandeep Seth of SethLaw PLLC in Texas is also
admitted pro hac vice. David R. Cooper of the Kansas firm Fisher, Patterson, Sayler & Smith,
LLP, is local counsel. Buether Joe & Counselors, LLC is handling this case on behalf of Lexos
on a contingent-fee basis. Mr. Buether and Mr. Kula appeared on behalf of Lexos at the in-
person claim construction hearing on September 21, 2023.1 According to Mr. Joe’s declaration,
“[i]n October 2024, the lead attorney in charge in this case[,] Eric Buether (and my law
partner)[,] passed away. I took his place as the lead attorney in charge of this case after his
passing.”2 Mr. Seth was permitted to appear as co-counsel in this matter pro hac vice on March
28, 2023.3 He has a small practice comprised of himself and a paralegal who also works as an
office manager. “[His] practice largely involves co-counseling with other firms.”4
Overstock filed a motion for summary judgment and two motions to exclude expert
testimony under Daubert on June 13, 2025.5 Relevant to the issues before the Court is
Overstock’s motion to exclude Samuel Russ, Ph.D., who Lexos designated as its technical expert
1 Doc. 98.
2 Doc. 217-5 ¶ 4. The Court notes that no withdrawal of appearance has been filed for Mr. Buether; he still
appears as an active attorney on the docket more than one year after his death.
3 Doc. 51. The Court is aware that Lexos has filed many other lawsuits around the country to enforce the
same patents at issue in this case. A short review of those cases on Westlaw indicates that the Buether law firm and
Mr. Seth together represent Lexos in several of those cases too. See, e.g., Lexos Media IP, LLC v. eBay Inc., 722 F.
Supp. 3d 1042 (N.D. Cal. 2024); Lexos Media IP, LLC v. Amazon.com, Inc., No. 2:22-CV-00169-JRG, 2023 WL
5723642 (E.D. Tex. Sept. 5, 2023); Lexos Media IP LLC v. MSC Indus. Direct Co., No. 3:22-CV-1736-X, 2023 WL
11967299 (N.D. Tex. Sept. 6, 2023).
4 Doc. 217-1 ¶ 13.
5 Docs. 174, 175, 177.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 2 of 36
3
in this case. In his opening report, Dr. Russ incorrectly identified the stipulated claim
construction for “cursor image” as “a movable image on a display screen whose position can be
controlled through a user interface.”6 In other words, he failed to include the second part of the
claim construction: “. . . and that indicates where user input can be received.”7 Overstock’s
motion to exclude Dr. Russ’s testimony argues that because he considered the incorrect claim
construction for “cursor image,” his opinion should be excluded.8 And as part of its motion for
summary judgment, Overstock argues that without Dr. Russ’s technical opinion, Lexos cannot
show as a matter of law that Overstock infringed.9
Lexos responded to Overstock’s motions on time.10 Counsel neither sought, nor
apparently discussed, seeking an extension of time. In their declarations submitted in response to
the Court’s Order to Show Cause, counsel broke down their division of labor on the responses to
Overstock’s June 13 motions. Mr. Kula worked on the response to Overstock’s other motion to
exclude, pertaining to Lexos’s damages expert, Justin Blok. Mr. Seth “was assigned the ultimate
responsibility to draft, review, finalize, and ensure proper filing of both” the summary-judgment
response and the response to Overstock’s motion to exclude Dr. Russ.11 Mr. Doell worked on
“the initial drafting of some sections”12 of the response to Overstock’s summary-judgment
motion, but Mr. Seth “was primarily responsible for drafting” 13 the response to Overstock’s
6 Doc. 174-2 at 4.
7 See Doc. 92-1 at 8, Doc. 99 at 6.
8 Doc. 174 at 8–14.
9 Doc. 177 at 31–35.
10 Docs. 190, 193, 194 (superseding Doc. 192).
11 Doc. 217-5 ¶ 8.
12 Doc. 217-3 ¶ 5.
13 Doc. 217-1 ¶ 5.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 3 of 36
4
motion to exclude Dr. Russ’s report and testimony and the response to the summary-judgment
motion. Mr. Joe, “never reviewed or looked at the documents at issue prior to their filing with
the Court.”14
Mr. Cooper did not research or draft any of these filings. He received the Texas
attorneys’ drafts on July 7, 2025, the same day they were due. He read them in their entirety and
generally understood the arguments being made therein. “[N]one of the quotes or citations drew
[his] attention, so [he] did not cite check either document. [He] instead relied upon Mr. Doell
and Mr. Seth, whom [he] understood to be the drafters of the documents, for the accuracy and
authenticity of the citations in the documents.”15
B. 	Incorrect Citations, Quotations, and Statements of Law
The response briefs to Overstock’s motion to exclude Dr. Russ and motion for summary
judgment contain multiple problematic legal citations and statements of authority that were
created through the use of generative AI, which were not checked and confirmed before filing.
First, and primarily, Plaintiff’s response to the motion to exclude Dr. Russ (Doc. 194) contains a
litany of problems: (1) nonexistent quotations; (2) nonexistent and incorrect citations; and (3)
misrepresentations about cited authority. Overstock submitted a reply brief disclosing the
defective citations, quotations, and statements of authority,16 and Mr. Seth admits that he used
generative AI to help research and draft the brief without confirming the accuracy of authority it
produced.17 Second, Part III.D of Plaintiff’s Response to Defendant’s Motion for Summary
14 Doc. 217-5 ¶ 7.
15 Doc. 217-2 ¶ 13.
16 Doc. 200-1 (providing a table of fabricated and misleading case citations in Plaintiff’s response brief to
the motion to exclude Dr. Russ).
17 Doc. 217-1 ¶¶ 5–9; Doc. 207 at 3 n.1 (admitting that Plaintiff’s response to the summary judgment
motion also contains “two paragraphs with the same mistake which were carried over concerning the Cordis,
Baldwin, and AstraZeneca cases” and requesting that “the Court strike these citations” (citations omitted)).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 4 of 36
5
Judgment (Doc. 193) addresses Overstock’s argument that Dr. Russ’s failure to incorporate the
correct claim construction for “cursor image” precludes its infringement theory. As part of that
argument, Lexos included several of the same defects from the response to Overstock’s motion
to exclude Dr. Russ.
The Daubert response, Doc. 194, contains the following erroneous citations, quotations,
or statements of authority:
1. 	Lexos cited Flexuspine, Inc. v. Globus Medical, Inc., 879 F.3d 1369, 1375 (Fed.
Cir. 2018) and i4i Ltd. P’ship v. Microsoft Corp., 598 F.3d 831, 856–57 (Fed. Cir. 2010), aff’d,
564 U.S. 91 (2011), for the proposition that “the Federal Circuit has cautioned against striking
expert opinions for immaterial or correctable errors in claim construction, particularly where the
overall analysis is sound and does not mislead the jury.” 18
Flexuspine contains no discussion of expert admissibility on the cited page number. On a
later page, the Federal Circuit considered a challenge to the district court’s ruling of
noninfringement on the basis that there was no evidence to support the second portion of the
plaintiff’s claim construction because the expert did not discuss it.19 But the court found no error
with the district court’s order and noted that the plaintiff “does not cite to any evidence that
might satisfy the second requirement of the claim construction.” 20 This case does not support the
proposition for which it was cited.
In Microsoft, the Federal Circuit considered the district court’s decision to admit expert
testimony on the appropriate damages calculation and to support the degree of infringement
through survey evidence. The court concluded that the damages expert’s royalty calculations
18 Doc. 194 at 6.
19 879 F.3d at 1377.
20 Id.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 5 of 36
6
were based on a sound methodology, and that the defendant’s objections went to the weight,
rather than the admissibility of the evidence.21 The court also considered a challenge to survey
evidence that was offered to support the degree of infringement.22 The defendant challenged the
survey’s methodology and the data used to extrapolate damage calculations from the survey, but
the court determined that the survey met the minimum standards required for relevance and
reliability.23 And the court found that the challenges were best addressed through cross-
examination and the presentation of contrary evidence.24 Neither challenge to the expert
evidence in this case related to an incomplete claim construction. This case does not support the
proposition for which it was cited.
2. 	Lexos cited United States v. Rodriguez-Felix, 450 F.3d 1117, 1123 (10th Cir.
2006) for the proposition that Tenth Circuit case law recognizes a preference for “addressing
admissibility disputes that go to weight rather than fundamental reliability [with] ‘[v]igorous
cross-examination, presentation of contrary evidence, and careful instruction on the burden of
proof,’ not wholesale exclusion.”25 This language is not in Rodriguez-Felix; it is in Daubert.26
Moreover, Rodriguez-Felix, a criminal case, addressed the specific issue of expert
testimony on the reliability of eyewitness identification.27 The Tenth Circuit determined that the
district court did not abuse its discretion in excluding a defendant’s expert testimony about the
reliability of eyewitness testimony, largely because such testimony was unlikely to assist the
21 Microsoft, 598 F.3d at 854.
22 Id. at 855–56.
23 Id. at 856.
24 Id.
25 Doc. 194 at 6.
26 Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 596 (1993).
27 Rodriguez-Felix, 450 F.3d at 1123–27.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 6 of 36
7
jury.28 Instead, the court explained that “skillful cross-examination provides an equally, if not
more, effective tool for testing the reliability of an eyewitness at trial.”29 Although the statement
of law by Lexos is not incorrect,30 this particular case neither includes the quotation attributed to
it nor discusses the court’s preference for testing expert testimony that goes to the weight of the
evidence through cross-examination, contrary evidence, and jury instructions.
3. 	Lexos twice cited Hockett v. City of Topeka, No. 19-4037-DDC, 2020 WL
6796766, at *3 (D. Kan. Nov. 19, 2020). It cited this case first for the proposition that
“[e]xclusion is a ‘drastic remedy’ and is only warranted where the report is irreparably unreliable
or prejudicially misleading.”31 Second, it cited Hockett to argue that inadvertent use of an
incomplete claim construction by an expert does not require exclusion, but instead cross-
examination and supplementation.32 The second citation includes an explanatory parenthetical
with a quotation from the case: “The exclusion of evidence is an extreme sanction, and courts
should prefer less severe remedies, particularly where the error appears inadvertent or can be
cured without prejudice.”33 Neither this case nor the quotation exists.34
4. 	Lexos cited AVM Techs., LLC v. Intel Corp., 927 F.3d 1364, 1370–71 (Fed. Cir.
2019) for the proposition that “deficiencies in a report that go to weight, not admissibility, should
28 Id. at 1125.
29 Id. (emphasis added).
30 See Daubert, 509 U.S. at 596.
31 Doc. 194 at 6.
32 Id. at 13.
33 Id.
34 This case number is a Social Security appeal, Russell v. Comm’r of Soc. Sec. Admin. The undersigned,
not District Judge Daniel D. Crabtree, presided over that case. There is no case against the City of Topeka in the
Court’s CM/ECF database by a plaintiff with the last name of Hockett. Nor can the Court locate this case in the
Westlaw database in any other federal jurisdiction.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 7 of 36
8
be resolved by cross-examination.”35 This citation is incorrect. The correct citation is AVM
Techs., LLC v. Intel Corp., 927 F. Supp. 2d 139 (D. Del. 2013). This case involved motions to
exclude two expert witnesses in a patent case: (1) an expert witness on royalty damages who
relied on a single settlement agreement on a different patent; and (2) the patent inventor’s
testimony about damages. The court determined that neither expert offered reliable testimony
under Daubert and granted motions to exclude them both.36 This case does not support the
proposition for which it was cited, even if the citation was correct.
5. 	Lexos quoted from and relied on Liquid Dynamics Corp. v. Vaughan Co., 449
F.3d 1209, 1224 (Fed. Cir. 2006).37 There are multiple problems with this authority. First,
Lexos provided the following quotation from this case in an explanatory parenthetical: “Expert
testimony should not be excluded simply because the expert applied an incorrect claim
construction, so long as the expert’s analysis can be understood and evaluated in light of the
court’s proper construction.”38 This quotation does not exist.
Second, Lexos cited this case for the proposition that “[t]he Federal Circuit has
consistently held that expert opinions based on even partially incorrect constructions may be
admissible where the expert’s methodology remains sound.”39 But Liquid Dynamics does not
support this proposition. To be sure, the court determined in that case that the defendant’s
challenge to certain parameters used by the expert in his analysis went to the weight and not the
admissibility of the opinion.40 But that ruling was not based on an incorrect claim construction
35 Doc. 194 at 6.
36 Intel Corp., 927 F. Supp. 2d at 141.
37 Doc. 194 at 7.
38 Id.
39 Id.
40 Liquid Dynamics, 449 F.3d at 1220–21.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 8 of 36
9
used by the expert. 	The Federal Circuit separately ruled as follows on the issue of the expert’s
reliance on an incorrect claim construction:
The district court excluded the expert opinion evidence [on
enablement] as irrelevant because it was based on an
impermissible claim construction and relied heavily on Figure 6 of
the patent specification that shows perfect helical flow instead of
substantial helical flow. Furthermore, the court found that the
evidence could prejudice and confuse the jury. Since the
enablement inquiry necessarily depends on an interpretation of the
claims, we conclude that the district court did not abuse its
discretion in excluding the expert’s testimony pertaining to
enablement.41
6. 	Lexos cited two cases for the proposition that “[t]he Federal Circuit has
consistently held that ‘technical flaws in an expert’s methodology or assumptions go to the
weight of the evidence rather than its admissibility.’”42 Immediately following this quotation,
Lexos cited Apple Inc. v. Motorola, Inc., 757 F.3d 1286, 1314 (Fed. Cir. 2014), but this
quotation does not appear in Apple, Inc. Lexos next cited Liquid Dynamics. The quote is not in
that case either, and, as described above, Liquid Dynamics holds contrary to Lexos’s statement of
law with respect to the relevance of an expert opinion based on an incorrect claim construction.43
7. 	Lexos cited two Tenth Circuit decisions for the proposition that under that court’s
standards, “exclusion is a drastic remedy reserved for fundamentally unreliable expert analysis,
not for technical deficiencies that can be addressed through cross-examination”: Rodriguez-Felix
and Goebel v. Denver & Rio Grande W. R.R., 215 F.3d 1083, 1088 (10th Cir. 2000).44 As
stated above, Rodriguez-Felix addresses the reliability of expert testimony on eyewitness
41 Id. at 1224 n.2.
42 Doc. 194 at 8.
43 Liquid Dynamics, 449 F.3d at 1224 n.2.
44 Doc. 194 at 9.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 9 of 36
10
credibility and excluded the expert’s opinion in that case. 45 And Goebel does not contain any
language about exclusion being a drastic remedy; it reversed the district court’s decision to admit
an expert report because it had not made sufficient findings under Daubert to support admitting
the expert testimony; it remanded for the district court to put more explicit findings on the record
for review.46 Neither case discussed nor applied Daubert in the context of a “technical
deficiency.”
8. 	Lexos explained the Federal Circuit’s decision in Cordis Corp. v. Medtronic Ave,
Inc., 511 F.3d 1157, 1174–75 (Fed. Cir. 2008) as follows:
[T]he Federal Circuit considered the phrase “expandable to a
diameter.” The Court held that this language did not limit the
structure to only that diameter, but rather described a capability or
result—not an exclusive state. Likewise, “modify . . . to [an
image] in the shape and appearance of [another image]” does not
demand exclusive substitution. Rather, it permits the modified
cursor to include or incorporate the specific image’s shape and
appearance, while retaining or overlaying other visual features.47
These quotations do not appear in Cordis. Almost as troubling, the patents that are discussed in
Cordis do not appear to contain any of the language quoted above.
9. 	Lexos included a citation to Baldwin Graphic Systems, Inc. v. Siebert, Inc., 512
F.3d 1338, 1342 (Fed. Cir. 2008) with the following quotation in an explanatory parenthetical:
“The use of ‘comprising’. . . renders the claim open-ended—meaning that the claim does not
exclude additional elements or steps.”48 This quotation does not exist.
45 450 F.3d 1117, 1123–27 (10th Cir. 2006).
46 Goebel, 215 F.3d at 1088–89.
47 Doc. 194 at 12.
48 Id.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 10 of 36
11
10. 	Lexos included a citation to AstraZeneca AB v. Mutual Pharm. Co., 384 F.3d
1333, 1339 (Fed. Cir. 2004) with the following quotation in an explanatory parenthetical: “The
open-ended term ‘comprising’ permits the inclusion of other steps or components.”49 This
quotation does not exist.
11. 	Lexos purported to quote the earlier-cited Microsoft Corp. case as follows: “[T]he
question of whether the expert is credible or whether his theories are correct given the partial
reliance on an incorrect claim construction is for the jury to decide after cross-examination.”50
This quotation does not exist.
Part III.D.1 of the summary-judgment response repeats the incorrect quotations from
Cordis,51 Baldwin Graphic Systems, Inc.,52 and AstraZeneca AB.53
C. 	Motion for Leave to Correct
In the reply to the response on the motion to exclude Dr. Russ, Overstock pointed out the
citation errors in Lexos’s opposition brief.54 Eight days later, Lexos filed a motion for leave to
file a corrected response brief.55 Lexos stated that after reading Overstock’s reply, it reviewed
the opposition brief and “discovered that certain case citations and quotations could not be
substantiated.” 56 The motion admitted that Mr. Seth used AI tools in drafting the opposition and
that he did not verify the citations before filing. In a footnote, the motion acknowledged that the
49 Id.
50 Doc. 194 at 13–14 (quoting I4i Ltd. P’ship v. Microsoft Corp., 598 F.3d 831, 854 (Fed. Cir. 2010), aff’d,
564 U.S. 91 (2011)).
51 Doc. 193 at 33–34.
52 Id. at 34.
53 Id.
54 Doc. 200 at 6 & n.1; Doc. 200-1.
55 Doc. 207.
56 Id. ¶ 3.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 11 of 36
12
summary-judgment response also contains “two paragraphs with the same mistake which were
carried over concerning the Cordis, Baldwin, and AstraZeneca cases” and requested that “the
Court strike these citations.”57
Mr. Seth submitted a declaration in conjunction with the reply in support of the motion
for leave to correct.58 No other attorney for Lexos did so. In that declaration, Mr. Seth
acknowledged the errors in Docs. 193 and 194, admitted that they were derived from his research
using generative AI that was not independently verified prior to submission, admitted that he
“authored the majority of the document,” and stated that he regrets the “oversight” and has put
into place “stricter internal review procedures to ensure that all future filings are fully sourced,
accurate, and independently verified.”59 Mr. Seth indicated in this declaration that at the time of
filing, he was struggling with multiple family health emergencies; his mother and aunt were both
hospitalized and ultimately passed away in July and early August. He indicated that “[t]hese
overlapping personal situations weighed heavily on [his] time, attention and emotional state
during the time of original filing,”60 and although he acknowledged that this does not excuse his
conduct, he claimed that it was unprecedented and that he did not intend to file the document
without first verifying the citations.
The proposed corrected response brief would have: (1) deleted the nonexistent citations
and quotations; (2) corrected the incorrect citations; (3) deleted the offending citations from the
table of authorities; (4) added one new citation; and (5) added text to shore up Lexos’s
arguments.
57 Id. at 3 n.1.
58 Doc. 211-1.
59 Id. ¶¶ 2, 3, 9.
60 Id. ¶ 7.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 12 of 36
13
Because there was nothing in the docket text that pointed this Court to an allegation of
sanctionable conduct, it was not alerted to this issue until it reviewed the motion to correct in
conjunction with the substantive motions several months later. On December 15, 2025, upon
review of the relevant filings, the Court denied Lexos’s motion for leave to correct to the extent
it sought to shore up its substantive research with new or corrected authority, or with new
arguments.61 The Court found that the prejudice to Overstock cannot be overstated, that the
errors have wasted judicial resources, and that the need to correct was well within the control of
Plaintiff’s counsel. The Court acknowledged Mr. Seth’s admission that he used generative AI
without validation, and his remorse, but it determined that a “do-over” was not warranted.
Instead, the Court denied the motion to file the proposed corrected brief, struck Plaintiff’s
response brief filed as Doc. 194, and granted Plaintiff’s motion to correct with a new filing that
removed the fake and incorrect citations, quotations, and misstated authority from the document.
Lexos timely filed this corrected document on December 19, 2025.62
Also on December 15, the Court entered an Order to Show Cause to all counsel of record
for Lexos. The Court was not satisfied with Mr. Seth’s declaration that was submitted along
with the reply brief on the motion to correct, given that four other attorneys’ names are listed on
the signature blocks of the offending documents. The Court directed each of the attorneys on the
signature blocks of Docs. 193 and 194 to show cause in writing, under penalty of perjury, why
they should not be sanctioned under Rule 11 and referred to the disciplinary panel of this Court
and to disciplinary administrators in the jurisdictions where they are licensed for submitting
briefs containing nonexistent or misrepresented authority. The Court put counsel on notice that
61 Doc. 214.
62 Doc. 216.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 13 of 36
14
failure to show cause may result in sanctions under Rule 11, including monetary sanctions,
attorneys’ fees, revocation of pro hac vice privileges, continuing legal education requirements,
and reporting obligations in their other cases and to the judges whose decisions were misquoted,
misrepresented, or made up. 63 Counsel have now filed their declarations. The following
findings of each attorney’s role are based on these declarations.
D. 	Lexos Attorneys’ Roles in Submitting Docs. 193 and 194
1. 	Mr. Seth
Mr. Seth has practiced law for over thirty years. He makes clear in his second declaration
that he alone used generative AI to insert the problematic citations, quotations, and discussions
of authority into the offending briefs. After writing an initial draft of the Daubert response on
June 28, 2025, Mr. Seth used ChatGPT “as a shortcut to find 10th [C]ircuit and Federal Circuit
case law consistent with the facts of the case.”64 That first search provided him with the
quotations and citations for the mistakes in items 3–5 and 11 above. He admits that he did not
check these citations, quotations, or statements of authority for accuracy. He then circulated the
draft to the “litigation team.”65
Mr. Seth then wrote a second draft that “substantially expanded the factual basis of our
argument.”66 Again, he queried ChatGPT to find additional case law. Part of that query was:
“[t]aking the role of a judge, write an order that denies the motion to strike with caselaw support
for the proposition that where the expert report is criticized for inadvertently using an immaterial
63 The Court recognizes that attorneys’ fees are not an option for sanctions when it acts sua sponte, as it did
here. See Fed. R. Civ. P. 11(c)(5)(B). Therefore, notwithstanding the language in its Order to Show Cause, the
Court will not impose this sanction.
64 Doc. 217-1 ¶ 7.a.
65 Id. Mr. Seth does not specify who the litigation team included.
66 Id. ¶ 7.b.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 14 of 36
15
incomplete claim construction, the remedy is not to strike the entire report/testimony of that
expert.”67 Mr. Seth believed this would give him the best chance of finding on-point case law.
That query resulted in the mistakes in items 1, 2, 6, and 7 above. Mr. Seth does not address the
nonexistent quotations set forth in items 8–10 above.
At some point, Mr. Doell provided minor edits to Mr. Seth’s draft, but these were
exclusively to add exhibit numbers and make small typographical suggestions. As the junior
attorney working on this case, he was not instructed to make substantive revisions or second
guess the senior co-counsel’s work. Mr. Doell made these minor edits and executed a
declaration for the exhibits.68
Mr. Seth contends that “neither co-counsel nor Lexos were aware of” his use of
generative AI, and that “Lexos was not billed for any of [his] time in the drafting of the brief
(and the portion of Document 193) containing the erroneous citations or for any subsequent time
in connection with their removal.”69
Mr. Seth contends that his family’s emergency medical situation led him to reach for
ChatGPT to help draft the Daubert response and that this was the first and only time he used it.
He claims to have been “untrained” on how to use it and a “novice.”70 He insists that using
ChatGPT is not his normal practice for research and that he made an isolated poor decision due
to the strain of end-of-life caregiving for his mother and aunt. Because his focus was elsewhere,
he failed to “double-check for accuracy.”71 He states that only now is he “fully aware of ‘AI
67 Id.
68 See Doc. 194-1.
69 Doc. 217-1 ¶ 6.
70 Id. ¶ 8.
71 Id. ¶ 10.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 15 of 36
16
hallucinations’ and take[s] full and sole responsibility for the resulting misinformation to the
Court.”72 At the same time, Mr. Seth states that he “inten[ded] at the time to check [the cites]
prior to any filing with the Court, but [he] failed to do so either because [he] thought [he] had
already done so, or just because of [his] scrambled state of mind.”73
2. 	Buether Joe and Counselors, LLC Attorneys
Mr. Joe is the lead attorney in this matter and the managing member of the Buether Joe
law firm. Mr. Joe played no role in drafting or compiling the defective citations in Docs. 193
and 194. In fact, he attests that he never reviewed or looked at either document prior to filing.
The Buether Joe law firm has a policy of not allowing the use of AI platforms by members of the
firm. He provides no information about how this policy is enforced.
Mr. Kula is a senior counsel at the Buether firm. He played no role in drafting or
compiling the defective citations in Docs. 193 and 194. He admits that he “never reviewed either
document at issue prior to them being filed with the Court.”74 Instead, Mr. Kula worked on
responding to the other Daubert motion that sought to exclude Lexos’s damages expert. When
Docs. 193 and 194 were filed, Mr. Kula was on a family vacation out of the country. Mr. Kula
understood that Mr. Doell was primarily responsible for drafting “limited sections” of the
summary-judgment response, but that Mr. Seth had “overall responsibility to research, draft,
review, finalize, and ensure the proper filing of both documents at issue.”75
Mr. Doell is an associate attorney at the Buether Joe law firm. As stated above, he
primarily drafted several sections of the summary-judgment response, which did not include
72 Id. ¶ 14.
73 Id. ¶ 9.
74 Doc. 217-4 ¶ 6.
75 Id. ¶¶ 11, 12.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 16 of 36
17
section III.D.1. Mr. Doell attests that his portion of the brief did not utilize AI; he understood
and abided by his firm’s policy prohibiting the use of AI platforms in writing, drafting, or
creating documents to be filed in court. Despite reading over and editing Mr. Seth’s work, he
was not aware, nor did he have reason to be aware, that Mr. Seth used AI in drafting the Daubert
response. He makes clear that in his subordinate role as an associate attorney working with co-
counsel who has 30 years of patent law experience, he “did not provide, nor intended to provide,
nor was expected to provide a substantive review of the caselaw or arguments in Mr. Seth’s draft
of Document 194.”76
3. 	Mr. Cooper
Mr. Cooper is local counsel. In that capacity, he signed Docs. 193 and 194 and filed
them on behalf of Lexos. Mr. Cooper is a partner at the Fisher Patterson law firm, and has
appeared in this Court many times over the last 25 years. He was aware that this Court’s local
rule requires that “[a]ll pleadings or other papers signed by an attorney admitted pro hac vice
must also be signed by a member of the bar of this court in good standing, who must participate
meaningfully in the preparation and trial of the case or proceedings to the extent the court
requires.”77
As already stated, Mr. Cooper received Docs. 193 and 194 on the day that they were due.
In a past filing in this case, Mr. Cooper “noticed an incomplete citation in a draft before filing,”
which he corrected.78 On July 7, 2025, he read the documents provided to him by the Texas
attorneys, but because none of the citations or arguments looked suspicious to him, he did not
cite check them. He had no knowledge that Mr. Seth used generative AI to compile his research,
76 Doc. 217-3 ¶ 11.
77 D. Kan. Rule 83.5.4(b).
78 Doc. 217-2 ¶ 14.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 17 of 36
18
or that he included defective citations and statements of authority in the briefs. He relied on Mr.
Doell and Mr. Seth to include accurate and authentic legal authority in the documents. Mr.
Cooper approved affixing his signature and then filed the offending documents in the Court’s
CM/ECF database.
Mr. Cooper apologizes to the Court and sets forth steps his firm has taken since the Court
issued its Order to Show Cause to ensure that this grievous error does not occur again. Fisher
Patterson has adopted a formal policy prohibiting the use of generative AI platforms without the
consent of the client and the express permission of the supervising partner or attorney on the
case. All documents produced using generative AI will be required to be independently verified
by the author as being true and accurate. The firm is also developing a policy that will extend
this policy to its attorneys’ practice as local or co-counsel, requiring local counsel to
independently cite-check and verify the accuracy of all citations.
II. 	Legal Standard
Under Rule 11(b):
By presenting to the court a pleading, written motion, or other
paper—whether by signing, filing, submitting, or later advocating
it—an attorney or unrepresented party certifies that to the best of
the person’s knowledge, information, and belief, formed after an
inquiry reasonable under the circumstances:
. . . .
(2) 	the claims, defenses, and other legal contentions are
warranted by existing law or by a nonfrivolous argument
for extending, modifying, or reversing existing law or for
establishing new law . . . .
As the Supreme Court has explained, “the central purpose of Rule 11 is to deter baseless filings
in district court and thus, consistent with the Rules Enabling Act’s grant of authority, streamline
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 18 of 36
19
the administration and procedure of the federal courts.”79 The rule “imposes . . . an affirmative
duty to conduct a reasonable inquiry into the facts and the law before filing, and . . . the
applicable standard is one of reasonableness under the circumstances.”80 “The award of Rule 11
sanctions involves two steps. The district court first must find that a pleading violates Rule 11.
The second step is for the district court to impose an appropriate sanction.”81
Where, as here, the Court invokes Rule 11 on its own initiative, it must order “an
attorney, law firm, or party to show cause why conduct specifically described in the order has not
violated Rule 11(b).”82 The Court met this requirement when it issued its Order to Show Cause
on December 15, 2025, giving Lexos’s counsel the opportunity to show cause in writing why
they should not be sanctioned for signing Docs. 193 and 194, which contain nonexistent citations
and quotations, and misstatements of legal authority, in contravention of Rule 11(b)(2).
III. 	Violation: Rule 11(b)
The Court now turns to the first step of the Rule 11 analysis: whether the Lexos attorneys
submitted a pleading that violates Rule 11(b). In determining whether an attorney’s conduct
violates Rule 11, the Court employs an “objective reasonableness” standard.83 The question is,
“whether a reasonable attorney admitted to practice before [this] court would file such a
document.”84 The Advisory Committee’s note to the 1993 Rule 11 Amendments explains that
the rule “require[s] litigants to ‘stop-and-think’ before initially making legal or factual
79 Cooter & Gell v. Hartmarx Corp., 496 U.S. 384, 393 (1990).
80 Bus. Guides, Inc. v. Chromatic Commc’ns Enters., Inc., 498 U.S. 533, 551 (1991).
81 Collins v. Daniels, 916 F.3d 1302, 1319 (10th Cir. 2019) (citation modified).
82 Fed. R. Civ. P. 11(c)(3).
83 Predator Int’l, Inc. v. Gamo Outdoor USA, Inc., 793 F.3d 1177, 1182 (10th Cir. 2015).
84 Id. (quoting Adamson v. Bowen, 855 F.2d 668, 673 (10th Cir. 1988)).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 19 of 36
20
contentions.” 85 And this duty is nondelegable.86 Thus, when an attorney signs a filing, that
attorney pledges that he or she conducted a reasonable inquiry into the existing law stated
therein.87 “[B]lind reliance on another attorney can be an improper delegation of this duty and a
violation of Rule 11.”88
There is no question that citing to a nonexistent case, attributing a nonexistent quotation
to an existing case, and misstating the law violates Rule 11(b).89 Such reliance does not
constitute “existing law” and certainly does not provide to the Court a nonfrivolous argument for
extending, modifying, or reversing existing law or for establishing new law. As the Tenth
Circuit recently noted, “there is nothing inherently problematic with the use of GenAI in the
practice of law.”90 Instead, the violation here is due to the failure to verify that the cases
generated by ChatGPT actually exist and confirm that they stand for the propositions for which
they are cited.91 Because there is no dispute that all five Lexos attorneys signed both documents
that included these errors, and they admit that not one of them verified that the case law in those
85 Fed. R. Civ. P. 11 advisory committee’s note to 1993 amendment.
86 Pavelic & LeFlore v. Marvel Ent. Grp., 493 U.S. 120, 126–27 (1989).
87 Bus. Guides, Inc. v. Chromatic Commc’ns Enters., Inc., 498 U.S. 533, 542 (1991) (“A signature certifies
to the court that the signer has read the document, has conducted a reasonable inquiry into the facts and the law and
is satisfied that the document is well grounded in both, and is acting without any improper motive.”).
88 Wadsworth v. Walmart Inc., 348 F.R.D. 489, 495–96 (D. Wyo. 2025); see also Dehghani v. Castro, 782
F. Supp. 3d 1051, 1058 (D.N.M. 2025) (finding a Rule 11(b) violation where attorney “outsourced his duties to
another attorney and failed to adequately review that attorney’s work-product and ensure its accuracy before putting
his own name on it and filing it with th[e] Court”).
89 See, e.g., Park v. Kim, 91 F.4th 610, 615 (2d Cir. 2024) (“At the very least, the duties imposed by Rule
11 require that attorneys read, and thereby confirm the existence and validity of, the legal authorities on which they
rely.”); Mata v. Avianca, Inc., 678 F. Supp. 3d 445, 461 (S.D.N.Y. 2023) (“A fake opinion is not ‘existing law’ and
citation to a fake opinion does not provide a non-frivolous ground for extending, modifying, or reversing existing
law, or for establishing new law.”); Coomer v. Lindell, No. 22-CV-01129, 2025 WL 1865282, at *3 (D. Colo. July
7, 2025) (“Courts across the country—both within the United States Court of Appeals for the Tenth Circuit . . . and
outside of it—recognize that Rule 11 applies to the use of artificial intelligence.”).
90 Moore v. City of Del City, No. 25-6002, 2025 WL 3471341, at *2 (10th Cir. Dec. 3, 2025).
91 See id.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 20 of 36
21
briefs actually exist and stand for the propositions for which they were cited, their conduct
violates Rule 11(b)(2).
All of the attorneys focus their declarations on their lack of bad faith in signing these
offending documents. But, as explained above, an attorney violates Rule 11 even when that
attorney does not subjectively intend to deceive the court.92 The standard is an objective one.
And the Court cannot find that any of these attorneys met this objective standard here. It is
undisputed that only Mr. Seth utilized ChatGPT and inserted the inaccurate legal authority from
those queries into the documents, and that none of the other attorneys were aware that he did so.
But all of the attorneys who signed the briefs had a nondelegable duty to conduct a reasonable
inquiry into the legal authority relied on in the briefs before signing their names.93 By signing
their names to Docs. 193 and 194, they “certified that each filing had been reviewed and verified
by human judgment,” yet they were not.94 As another district court in this circuit has noted
under similar circumstances, “ignorance does not excuse [this attorney’s] Rule 11(b) violations.
It was [the attorney’s] responsibility to ensure that the Brief, which he signed and filed, was
accurate.”95
92 See White v. Gen. Motors Corp., 908 F.2d 675, 680 (10th Cir. 1990) (“A good faith belief in the merit of
an argument is not sufficient; the attorney’s belief must also be in accord with what a reasonable, competent attorney
would believe under the circumstances.”); In re Allen, 248 F. App’x 874, 882 (10th Cir. 2007) (affirming district
court’s application of an objective Rule 11 standard when imposing sua sponte sanctions); Omniq Corp. v. Redlpr,
LLC, No. 19-CV-437, 2025 WL 525044, at *9 (D. Utah Feb. 18, 2025) (rejecting argument that sua sponte sanctions
must be decided under a subjective bad faith standard in the Tenth Circuit). But see Kyros L. P.C. v. World
Wrestling Ent., Inc., 78 F.4th 532, 543 (2d Cir. 2023) (“When a court initiates Rule 11 sanctions sua sponte and the
opportunity to correct or withdraw the challenged submission is unavailable, the court must make a finding of bad
faith on the part of the attorney before imposing the sanctions.”); United Nat. Ins. v. R&D Latex Corp., 242 F.3d
1102, 1115–16 (9th Cir. 2001) (explaining that sua sponte sanctions under Rule 11 should “ordinarily be imposed
only in situations that are akin to a contempt of court”).
93 Mattox v. Prod. Innovations Rsch., LLC, No. 6:24-CV-235-JAR, 2025 WL 3012828, at *5 (E.D. Okla.
Oct. 22, 2025) (“Responsibility under Rule 11 extends to all attorneys of record.”).
94 Id. at *4.
95 Dehghani v. Castro, 782 F. Supp. 3d 1051, 1059 (D.N.M. 2025).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 21 of 36
22
The Court finds that Mr. Doell was placed in an untenable position by his superiors. He
was directed to draft much of the summary-judgment response, not including the portion that
came from Mr. Seth’s faulty ChatGPT research. There is no indication that his own work
product utilized unverified research, and he attests under penalty of perjury that it did not. He
was also told to provide limited edits of Mr. Seth’s work, but not to conduct a substantive
review. The Court is mindful that this created a difficult power differential—Mr. Doell is the
associate attorney on this matter, and Mr. Seth was co-counsel bringing 30 years of experience
and no prior ethical lapses or violations to the table. Mr. Doell’s superiors at the Buether Joe
firm apparently had no role in reviewing these two documents or supervising his work. One of
them was out of the country on the date they were filed. Nonetheless, Mr. Doell played a role in
drafting and editing these documents, and under the objective standard that applies to Rule
11(b)(2), he did not make a reasonable inquiry into the law relied on by Mr. Seth before he
signed them. Had he done so, he would have discovered these many errors.
Mr. Joe and Mr. Kula did not even read the documents before they were filed. They
should have known that giving signature authority without review runs the risk of violating Rule
11. As another district court in the Tenth Circuit has explained, “[w]hen an attorney gives
another permission to sign on their behalf without reviewing the document, it is not only
actionable under Rule 11 but also a possible violation of a state’s ethical rule of competence.”96
Indeed, “no inquiry cannot be deemed objectively reasonable even if the reliance is placed in an
experienced attorney.”97 There was a simple way for these attorneys to avoid the risk of a
96 Wadsworth v. Walmart Inc., 348 F.R.D. 489, 496 (D. Wyo. 2025).
97 Id.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 22 of 36
23
violation: they did not have to sign the documents. By doing so, they placed their imprimatur on
the filings and are therefore subject to the consequences of that decision.98
Mr. Cooper, as local counsel, read the documents, but he failed to cite-check them,
despite having discovered a citation error in a previous brief submitted to him for filing in this
case.99 He admits relying on his co-counsel to properly cite the cases therein when he signed his
name to the documents. But by blindly relying on his co-counsel, he delegated a nondelegable
duty. Adherence to Rule 11 is particularly important in the context of local counsel. In this
district, counsel admitted pro hac vice may not directly file documents in CM/ECF. They are
required to use local counsel to do so. And local counsel gives those filings “a presumption of
legitimacy.”100 This is why the Court’s local rule requires pleadings prepared by pro hac vice
attorneys be signed by local counsel who is in good standing and that local counsel meaningfully
participate in the proceedings.101 Mr. Cooper abdicated this responsibility by not confirming the
accuracy of the authority in Docs. 193 and 194 before affixing his signature and filing them
under his own name.
There was simply no oversight provided by the many seasoned attorneys of record for
Lexos when responding to these motions, despite the fact that they each had a duty under Rule
11, not to mention the rules of professional responsibility, to ensure that the documents they
98 See Mattox, 2025 WL 3012828, at *8 (“When a lawyer lends his signature, or his silence, to filings
without inquiry, he converts inattention into representation and transforms friendship into liability.”).
99 See Doc. 217-2 ¶ 14 (“The process described in ¶ 13, supra, differs from the review and approval of at
least one earlier filing where I had noticed an incomplete citation in a draft before filing. I found the original case
and corrected the citation before filing the document.”).
100 Mattox, 2025 WL 3012828, at *8.
101 See D. Kan. Rule 83.5.4; Doc. 51 (Mr. Cooper’s signed pro hac vice application on behalf of Mr. Seth
agreeing to “participate meaningfully in the preparation and trial of this case.”).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 23 of 36
24
signed were based on existing law. As such, they all violated Rule 11 when they signed or
allowed their names to be added to the signature blocks of Docs. 193 and 194.
IV. 	Sanctions: Rule 11(c)
While the Rule 11(b) violation in this case is crystal clear to the Court, the appropriate
sanction for each attorney is more difficult. The rule dictates that any sanction imposed “must be
limited to what suffices to deter repetition of the conduct or comparable conduct by others
similarly situated.”102 And the Court must consider the relative fault of each attorney in
fashioning sanctions.103
This Court has already stricken Doc. 194 and granted leave to Lexos to file a corrected
document that removes the offending citations. As a practical matter, this removed all but three
case citations that counsel previously submitted in support of the motion, and all of the
substantively-specific authority Lexos previously used to persuade the Court not to grant the
motion to strike. 	Although the Court did not explicitly address the summary-judgment response
in its order on the motion for leave to correct, at this time it also strikes the offending authority
that was copied from the Daubert response into Part III.D of the summary-judgment response.
Beyond this consequence of counsel’s conduct, the Court discusses below the relative fault of
each attorney and the sanctions, if any, it imposes.
In considering the range of sanctions available, the Court cannot ignore the context in
which these violations occur. The sheer amount of case law that has erupted over the last few
years due to attorneys’ reliance on unverified generative AI research, often generating
102 Fed. R. Civ. P. 11(c)(4).
103 White v. Gen. Motors Corp., 908 F.2d 675, 685–86 (10th Cir. 1990).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 24 of 36
25
hallucinated legal authority, is staggering.104 There are too many sanctions orders at this point
by other courts to cite them all,105 and the problem has been reported in news sources.106 The
American Bar Association107 and many state bar associations have issued guidance. 108 Several
courts now either have notices, local rules, or standing orders that address the use of generative
AI by litigants.109
104 See Anna Conley, Understanding the Duty of Competence for Attorneys Using Generative AI, 27 N.C. J.
L. & Tech. 65, 75 & n.38 (2025) (stating that the author had located over 40 different cases involving attorneys
citing generative AI with nonexistent cases and citing to high profile examples); see also Evan Gorelick, Vigilante
Lawyers Expose the Rising Tide of A.I. Slop in Court Filings, N.Y. Times (Nov. 11, 2025), at B3 (referencing
lawyer and researcher Damien Charlotin’s attempt to create online databases tracking cases that deal with fabricated
cases generated by A.I. and noting at that time that he had documented 509 such cases); Tracking Federal Judge
Orders on Artificial Intelligence, Law360, https://www.law360.com/pulse/ai-tracker (last visited Jan. 27, 2026)
(aggregating 59 cases).
105 See, e.g., Johnson v. Dunn, 892 F. Supp. 3d 1241, 1266 (N.D. Ala. 2025) (collecting cases); Mattox v.
Prod. Innovations Rsch., LLC, No. 24-CV-235, 2025 WL 3012828, at *6–7 (E.D. Okla. Oct. 22, 2025) (same).
106 See, e.g., Benjamin Weiser, So, Have You Heard the One About the Lawyer Using A.I.?, N.Y. Times
(May 29, 2023), at A1; Sarah Martinson, Judges’ AI Orders Keep Trickling in as Fake Citations Persist, Law360
(July 21, 2025), https://www.law360.com/pulse/articles/2357542?; Daniel Wu, Lawyers Using AI Keep Citing Fake
Cases in Court. Judges Aren’t Happy, Washington Post (June 3, 2025),
https://www.washingtonpost.com/nation/2025/06/03/attorneys-court-ai-hallucinations-judges/.
107 A.B.A. Comm. on Ethics & Pro. Resp., Formal Op. 512 (2024).
108 See A.B.A. Task Force on Law & A.I., Addressing the Legal Challenges of AI, Year 2 Report on the
Impact of AI on the Practice of Law, at 47 (Dec. 2025) (compiling state bar ethics rules and guidance). Mr. Seth is
licensed in Texas and California. Both state bars provide guidance on attorneys’ use of AI. State Bar of Cal.
Standing Comm. on Pro. Resp. & Conduct, Practical Guidance for the Use of Generative Artificial Intelligence in
the Practice of Law (2023) (providing guiding principles regarding how the rules of professional responsibility
should be applied to lawyers’ use of generative AI, including that under Rules 1.1 and 1.3, “AI-generated outputs
can be used as a starting point but must be carefully scrutinized. They should be critically analyzed for accuracy and
bias, supplemented, and improved, if necessary. . . . A lawyer’s professional judgment cannot be delegated to
generative AI and remains the lawyer’s responsibility at all times”); Tex. Comm. on Pro. Ethics, Op. 705 (Feb.
2025) (“While there may be many ways that generative AI can assist in the practice of law and benefit lawyers and
clients alike, Texas lawyers must always be aware of the ethical issues that may arise in the use of generative AI.
Among many other issues, lawyers should acquire basic technological competence before using any generative AI
tool, should always ensure that the tool does not imperil confidential client information, should always verify the
accuracy of any responses received from a generative AI tool, and should not charge clients for the time ‘saved’ by
using a generative AI program.”).
109 See, e.g., United States District Court for the District of Connecticut, Notice to Counsel and Litigants
Regarding AI-Assisted Research, at https://www.ctd.uscourts.gov/; E.D. Tex. Admin. R. III(3)(m); In re Use of
Artificial Intelligence, No. 3:24-mc-104, slip op. (W.D.N.C. June 18, 2024), at
https://www.ncwd.uscourts.gov/sites/default/files/general-orders/AI%20Standing%20Order.pdf. See generally
RAILS, Analysis of AI Use in Courts, at https://rails.legal/resources/resource-ai-orders/ (providing table of court
rules, local rules, and judicial guidelines governing the use of AI through May 2025).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 25 of 36
26
A reasonably competent attorney filing documents in court should be aware of the
pronounced, well-publicized risks of using unverified generative AI for legal research and the
ethical obligations associated with signing a court filing without checking it for accuracy.
Nonetheless, all of the attorneys here disclaim such knowledge. Mr. Seth claims he was a novice
and did not appreciate the risk of case hallucinations. The Buether Joe attorneys suggest that
because their firm had a strict prohibition on the use of AI, they had no further obligation to
check their own work, and no reason to think that co-counsel, who was not covered by this
internal policy, used AI. Mr. Cooper relied on the Texas attorneys to cite check their work. The
Court finds that these attorneys’ collective failure to devise a system of checking their filings for
accuracy in the face of the well-publicized need to do so requires sanctions in order to deter
repetition of this conduct by them and by other attorneys who are similarly situated.
The Court has considered the many other cases where judges have sanctioned attorneys
for similar conduct. In most of them, courts have imposed monetary penalties ranging from as
high as $31,000, 110 to as low as $500. 111 In addition to the admonishment contained in this
Order, the Court imposes fines proportionate to the individual attorney’s violation in this case,
with the exception of Mr. Doell, as well as other targeted sanctions designed to ensure that these
attorneys and others similarly situated do not commit similar infractions in the future.
110 Lacey v. State Farm Gen. Ins., No. CV 24-5205, 2025 WL 1363069, at *5 (C.D. Cal. May 5, 2025)
(special master order sanctioning the plaintiff’s law firms, including ordering them to pay compensation to the
defense in the total amount of $31,000).
111 See, e.g., Wadsworth v. Walmart Inc., 348 F.R.D. 489, 495–96 (D. Wyo. 2025) ($3,000 fine for drafter;
$1,000 for other attorneys who signed); Cojom v. Roblen, LLC, No. 23-CV-1669, 2025 WL 3205930, at *4 (D.
Conn. Nov. 17, 2025) ($500 fine); Coomer v. Lindell, No. 22-1129, 2025 WL 1865282, at *8 (D. Colo. July 7,
2025) ($3,000 per attorney); Mattox v. Prod. Innovations Rsch., LLC, No. 6:24-CV-235-JAR, 2025 WL 3012828, at
*8 (E.D. Okla. Oct. 22, 2025) ($3,000 fine for drafter, $2,000 for counsel of record who did not sign, and $1,000 for
signer who did not review).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 26 of 36
27
A. 	Mr. Seth
There is no doubt on this record that Mr. Seth carries the most relative fault. He admits
that he used ChatGPT without verifying its research results, he admits to not telling his co-
counsel or the client that he did so, and he allowed his co-counsel to add their names to the
documents despite this knowledge. In fact, Mr. Seth asks the Court not to sanction any of the
other attorneys for the violation and states that he is solely to blame.
In fashioning an appropriate sanction for Mr. Seth, the Court is mindful of the difficult
family circumstances he was dealing with at the time of filing. This situation was so serious that
Mr. Seth states it had “a profound effect on [his] mental state at the time of this error that left
[him] distraught and somewhat distracted.”112 While the Court is sympathetic to Mr. Seth’s
family circumstances, it cannot ignore that there were obvious options open to him short of using
a publicly-available generative AI search tool that he had never used before to research his brief
without verifying the accuracy of that research. The Court cannot understand why he did not
share his situation, even generically, with one or more of his many co-counsel who could have
helped him. Mr. Kula handled the claim construction hearing and should have been in a position
to assist him on this Daubert issue. Another obvious option for Mr. Seth would have been to
seek an extension of time to file his response brief. Certainly caring for and attending to
extremely sick or dying family members would have constituted good cause for an extension of
time.113 Yet, he chose neither of these options. Instead, Mr. Seth chose to use ChatGPT, not
once but twice, and then failed to confirm that the very strong authority it generated was
accurate.
112 Doc. 217-1 ¶ 10.
113 See Fed. R. Civ. P. 6(b)(1)(A).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 27 of 36
28
Mr. Seth claims that this was the first and only time he used this tool that he admits he
did not understand. And despite using this tool for the first time while admittedly not
understanding it, he claims that he intended to but did not check the accuracy of its output. He
blames this oversight on his emotional state at the time. But the Court finds that, in addition to
his failure to verify the accuracy of his research, Mr. Seth’s admitted use of a generative AI tool
without understanding it is an aggravating circumstance that it must consider when imposing a
sanction.
The Court also takes into account Mr. Seth’s remorse and the fact that he attempted to
correct the defective pleading. Mr. Seth apologizes to the Court, his co-counsel, and opposing
counsel in his declarations, and assures the Court that he has learned his lesson. While the Court
is heartened by his remorse, several statements in Mr. Seth’s declaration leave it convinced that
further sanctions are necessary to deter future conduct both by Mr. Seth and by others similarly
situated.
First, Mr. Seth overemphasizes the fact that he did not deliberately intend to deceive the
Court.114 The Court recognizes that he did not set out to deliberately deceive the Court when he
chose to use ChatGPT to generate case citations, quotations, and statements of the law that he
then inserted into the briefs without checking them for accuracy. But if Mr. Seth was not aware
of the very real risk of case hallucinations resulting from unverified generative AI legal research,
he should have been. As referenced above, the number of cases, published and unpublished, that
resulted in Rule 11 sanctions for this very conduct has ballooned in recent years. And it has been
highly publicized. The states in which Mr. Seth is licensed have issued guidance. A reasonable,
114 Doc. 217-1 ¶¶ 12–14.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 28 of 36
29
competent attorney in Mr. Seth’s position should have been aware of the risks of using this
research tool.
Second, the Court is troubled by Mr. Seth’s assertion that he was merely seeking to
corroborate “the approximately 95% of humanly generated brief and substantive arguments
contained therein [that he] personally drafted based upon the facts in this case and not any AI
output.”115 This statement conveys that he does not appreciate the magnitude of his error. A
quick comparison of Mr. Seth’s original and corrected draft belies his claim that the original
brief was composed of 95% human-generated content.116 The correct brief is approximately two
pages shorter than the faulty one. The eleven misstatements of authority—all either nonexistent
or incorrect—pervaded Doc. 194.
Moreover, the only human-generated content in the original brief is either generic, or
made without citation to authority. The false quotations, citations, and misrepresented authority
were all highly specific to the issue raised in Overstock’s motion. And, as set forth in detail
earlier in this opinion, these statements of authority were quite strong. Mr. Seth did not merely
query ChatGPT for a few citations to add after his own text. He asked it to take on the role of
judge and provide him with authority and language opposing the motion. All citations in the
brief, other than the generic references to the Daubert standard, were generated by his unverified
use of ChatGPT. All of them were problematic. To deflect attention from this reckless conduct
by focusing on the other language in the brief is inconsistent with the remorse that Mr. Seth
expresses in other parts of his declaration. The Court instead weighs heavily the severity of the
115 Id. ¶ 14.
116 Compare Doc. 194, with Doc. 216.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 29 of 36
30
violation when fashioning Mr. Seth’s sanction—the repeated instances of hallucinated case
citations and quotations, and strong misstatements of authority.
In fashioning a sanction that focuses on deterrence, the Court is also mindful that Mr.
Seth has and continues to represent Lexos and others in the field of patent law in courts around
the country. Mr. Seth contends that much of his work is through contracts with co-counsel
arrangements similar to the one in this case—where he completes work as co-counsel on briefs
joined by other attorneys. Thus, his sanction must be designed to deter the violative conduct in
this and other cases.
Mr. Seth indicates that it is his “practice going forward to never use AI to identify case
law or quotes attributed to case law from AI,” and that his employees and independent
contractors will be advised the same.117 Respectfully, this is insufficient. As stated throughout
this Order, the Rule 11(b) violation was not necessarily counsel’s use of generative AI to conduct
legal research. The violation was (1) failing to understand the technology so that he could use it
in an informed way; and (2) failing to check the research it generated for accuracy.118 “As
attorneys transition to the world of AI, the duty to check their sources and make a reasonable
inquiry into existing law remains unchanged.” 119 Nothing about Mr. Seth’s stated policy
addresses the need for competence and verification. Thus, the Court directs Mr. Seth to
implement a more robust policy designed to deter any future instance of submitting unverified
authority in a filing. The Court will require him to submit to the Clerk for filing a certificate
117 Doc. 217-1 ¶ 11.
118 Whether Mr. Seth also violated the rules of professional responsibility is not before this Court. The
Court will defer to Mr. Seth’s state disciplinary authorities to make that determination.
119 Wadsworth v. Walmart Inc., 348 F.R.D. 489, 493 (D. Wyo. 2025).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 30 of 36
31
outlining specific internal procedures at his firm that he intends to impose that will ensure that
the legal authority contained in the firm’s work product is accurate.
With these considerations in mind, the Court imposes a monetary fine of $5,000 against
Mr. Seth and revokes his pro hac vice admission to this Court, given that several other competent
counsel appear in this matter on behalf of Lexos. The Court further directs Mr. Seth to self-
report to the state disciplinary authorities where he is licensed by providing them with a copy of
this Order, and to submit to the Clerk for filing by no later than February 28, 2026, a certificate
outlining the specific internal procedures at his firm that will ensure that the legal authority
contained in the firm’s work product is accurate.
B. 	Mr. Kula
As already discussed, Mr. Kula did not generate the defective citations, nor did he
contribute in any way to Docs. 193 and 194. Mr. Kula’s violation was signing a filing despite
failing to review it or otherwise ensuring that its contents were accurate. Mr. Kula instead
assigned an associate, Mr. Doell, to work on the majority of the summary-judgment brief, and to
assist Mr. Seth with both defective briefs. Mr. Kula left for a family vacation several days before
these briefs were filed.
Because he had no knowledge that Mr. Seth used unverified generative AI to help draft
Doc. 194, and inserted some of that defective authority into Doc. 193, and because he was only
responsible for handling the response to a different Daubert motion, Mr. Kula asks that sanctions
not be imposed. But as the Court has already explained, Mr. Kula’s lack of knowledge that Mr.
Seth used hallucinated authority in Doc. 194 does not absolve him of the Rule 11(b) violation. It
is his failure to review these briefs at all, despite affixing his name to the documents, that
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 31 of 36
32
constitutes sanctionable conduct; he violated his duty to conduct a reasonable inquiry into the
facts and the law before filing.
The Buether Joe attorneys appear to take the position in their declarations that because
the firm had a flat prohibition on the use of “AI platforms by firm personnel or attorneys for
writing, drafting, or creating documents to be filed with any court” and because they had no
reason to believe Mr. Seth used generative AI to draft Doc. 194, they should not be
sanctioned.120 Neither Mr. Kula nor Mr. Joe accept responsibility for signing the documents or
failing to supervise their preparation. Nor do they show any remorse for the fact that Docs. 193
and 194 were filed bearing their names with fabricated and misrepresented legal authority.
Given Mr. Kula’s abdication of his Rule 11 duties by affixing his name to documents that he
failed to review, and his failure to acknowledge this breach, the Court finds a sanction is
warranted. In addition to the public admonishment in this Order, the Court imposes a fine of
$3,000.
C. 	Mr. Joe
Like Mr. Kula, Mr. Joe states that he took no part in drafting or reviewing the defective
filings in this case, despite allowing his name to be affixed to the documents and despite being
the lead attorney in this matter. And, like Mr. Kula, Mr. Joe fails to acknowledge that he
breached a duty under Rule 11 by delegating his duty to conduct a reasonable inquiry into the
facts and law in Docs. 193 and 194 before signing them. For the same reasons the Court
sanctions Mr. Kula, the Court finds that a sanction is warranted against Mr. Joe in the form of
this public admonishment and a $3,000 fine.
120 Doc. 217-4 ¶ 15; Doc. 217-5 ¶ 11.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 32 of 36
33
In addition, given Mr. Joe’s role as the managing member of his law firm, the Court
imposes the additional sanction of ordering him to reconsider his firm’s internal policies
addressing the use of AI for generating filings in court. As the Court explained in ordering the
same for Mr. Seth, a mere prohibition on the use of AI platforms without any training,
enforcement, or verification procedure does not ensure the Court that the Buether Joe firm will
not be placed in a similar situation in the future, particularly if they submit briefs with co-counsel
from outside the firm who would not be subject to its policy. As the ABA’s guidance provides:
Model Rules 5.1 and 5.3 address the ethical duties of
lawyers charged with managerial and supervisory responsibilities
and set forth those lawyers’ responsibilities with regard to the firm,
subordinate lawyers, and nonlawyers. Managerial lawyers must
create effective measures to ensure that all lawyers in the firm
conform to the rules of professional conduct, and supervisory
lawyers must supervise subordinate lawyers and nonlawyer
assistants to ensure that subordinate lawyers and nonlawyer
assistants conform to the rules. These responsibilities have
implications for the use of [generative AI (“GAI”)] tools by
lawyers and nonlawyers.
Managerial lawyers must establish clear policies regarding
the law firm’s permissible use of GAI, and supervisory lawyers
must make reasonable efforts to ensure that the firm’s lawyers and
nonlawyers comply with their professional obligations when using
GAI tools. Supervisory obligations also include ensuring that
subordinate lawyers and nonlawyers are trained, including in the
ethical and practical use of the GAI tools relevant to their work as
well as on risks associated with relevant GAI use. Training could
include the basics of GAI technology, the capabilities and
limitations of the tools, ethical issues in use of GAI and best
practices for secure data handling, privacy, and confidentiality. 121
Thus, the Court directs Mr. Joe to cause the implementation of specific internal
procedures at his firm that will ensure that the legal authority contained in the firm’s future court
filings are accurate. The Court strongly encourages him to consider verification and training
121 A.B.A. Comm. on Ethics & Pro. Resp., Formal Op. 512, at 10 (2024) (footnotes omitted).
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 33 of 36
34
requirements for all members of his firm. He shall file by no later than February 28, 2026, a
certificate outlining these procedures.
D. 	Mr. Doell
As the Court has already discussed, as the most junior attorney on this matter without
supervisory authority, Mr. Doell was placed in a difficult position by his supervising attorneys.
He was apparently neither expected nor instructed to substantively check Mr. Seth’s work,
despite being asked to execute a declaration in support of the exhibits attached to it, and despite
being asked to conduct a technical edit. One of his supervising attorneys at the firm was out of
the country, and the other took no part in preparing, reviewing, or supervising the July 7, 2025
filings. The Court finds that the public admonishment of this Order is sufficient to deter any
similar future conduct by Mr. Doell and releases him from further sanctions. The remedial
sanctions that the Court imposes on Mr. Joe will apply to him as an associate at the firm and
should prevent him from being placed in this situation again.
E. 	Mr. Cooper
As local counsel, Mr. Cooper signed the defective filings. By doing so, he vouched for
the Texas attorneys in this matter. Although Mr. Cooper at least reviewed the filings before he
submitted them to the Court, he did not cite-check them. But the Court weighs heavily that,
unlike Mr. Kula and Mr. Joe, Mr. Cooper acknowledges that he had a duty to check the filings
for accuracy before signing and filing them under Rule 11 and that such duty was an objective
one; he expresses remorse for failing to meet this duty, and he sets forth in detail his and his
firm’s efforts to ensure that a similar infraction does not occur again.
The Fisher Patterson firm has adopted a formal policy (1) prohibiting the use of
generative AI platforms without the consent of the client and the express permission of the
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 34 of 36
35
responsible partner and, if different, the overseeing attorney, and (2) requiring that all documents
produced using generative AI be independently verified by the author as being true and accurate.
The firm is in the process of developing a policy that extends the formal AI policy to all
instances where an attorney in the firm is local counsel and sponsors other attorneys as pro hac
vice. In addition, Mr. Cooper has voluntarily sanctioned himself in the form of refraining from
serving as sponsoring or local counsel for pro hac vice attorneys for a period of 12 months.
Thus, the Court finds that the admonishment in this Order, the policies adopted by Mr.
Cooper’s firm, and a $1,000 fine are sufficient to deter Mr. Cooper and any other local counsel
who may be similarly situated from committing a similar infraction in the future.
IT IS THEREFORE ORDERED BY THE COURT that the following language is
stricken in Lexos’s response to Overstock’s summary-judgment motion: Doc. 193 at 38 and 39,
the entire paragraph under the heading “1. ‘Modify . . . to’ Does Not Imply Exclusivity” and the
citations on page 39 to Baldwin Graphic Systems, Inc. and AstraZeneca.
IT IS FURTHER ORDERED BY THE COURT that all five Lexos attorneys are
publicly admonished for violating Fed. R. Civ. P. 11(b) as set forth in this Order. The individual
attorneys are further sanctioned under Rule 11 as follows:
• 	Mr. Seth is fined $5,000, the Court revokes his pro hac vice admission to this
Court, he shall self-report to the state disciplinary authorities where he is licensed
by providing them with a copy of this Order, and he shall submit to the Clerk for
filing by no later than February 28, 2026, a certificate outlining specific internal
procedures to be imposed at his firm that will ensure that the legal authority
contained in the firm’s future court filings are accurate.
• 	Mr. Kula is fined $3,000.
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 35 of 36
36
• 	Mr. Joe is fined $3,000 and shall file, by no later than February 28, 2026, a
certificate outlining specific internal procedures to be imposed at the Buether Joe
law firm that are designed to ensure that the legal authority contained in the firm’s
future court filings are accurate.
• 	Mr. Cooper is fined $1,000.
• 	Mr. Doell is released from any further sanctions.
The fines imposed by this Order shall be made payable to the Court’s registry within 14
days of this Order.
IT IS SO ORDERED.
Dated: February 2, 2026
S/ Julie A. Robinson
JULIE A. ROBINSON
UNITED STATES DISTRICT JUDGE
Case 2:22-cv-02324-JAR-ADM Document 218 Filed 02/02/26 Page 36 of 36

Setzen Sie Ihre Recherche in ChatGPT oder Claude fort

Verbinden Sie Omnilex, um den Rechtskorpus über Ihren KI-Assistenten zu durchsuchen.